Phosphorylation and dephosphorylation are processes by which phosphate groups are added or removed from a target molecule, typically a protein. The process of reversible phosphorylation is a key feature of cellular regulation, including signal transduction, gene expression, cell cycle regulation, cytoskeletal regulation and apoptosis. See, e.g., PROTEIN PHOSPHORYLATION (Marks F. ed., 1996); Hunter, “Signaling—2000 and beyond,” Cell 100:113-127 (2000). Principally, two classes of enzymes (kinases and phosphatases) modulate reversible protein phosphorylation, adding phosphate groups and removing phosphate groups, respectively, from molecules. Phosphorylation reactions are key features of protein function, and thus phosphorylated proteins must be able to be identified if the proteome is to be fully understood; however, to date no practical methods exist for the systematic parallel analysis of the phosphorylation status of large sets of proteins involved in the regulatory circuitry of cells and tissues. See, Wilkins et al., Genetic Eng. Rev. 13:19 (1995).
Signal transduction is an example of a process involving protein phosphorylation that is critical for cellular regulation. After an extracellular stimulatory factor binds to its recognized cell surface receptor, signal transduction is initiated, often by a specific set of cellular protein kinases. These kinases subsequently phosphorylate the target molecule, resulting in an altered activity and a continued cellular response to the signal. See, e.g., Nishizuka, “Studies and perspectives of protein kinase C,” Science 233:305-312 (1986). It is not enough for researchers to simply identify whether a protein is a phosphorylated protein or not. It has become additionally essential for researchers to identify the sites of phosphorylation on proteins and to determine the stoichiometry of phosphorylation. Serine, threonine and tyrosine amino acid residues are the most common sites of phosphorylation in eukaryotic cells. See, e.g., Guy et al. “Analysis of Cellular Phosphoproteins by Two-Dimensional Gel Electrophoresis: Applications for Cell Signaling in Normal and Cancer Cells,” Electrophoresis 15:417-440 (1994). Thus, the focus for researchers in understanding protein phosphorylation events occurs at two levels. The first level of analysis requires a determination of whether a protein is a phosphoprotein, including identifying molecules responsible for phosphorylation, and the second level of analysis requires the identification of which amino acid is phosphorylated and how many amino acids are phosphorylated. The present invention provides materials and methods for both levels of analysis. The present invention also provides materials and methods for analysis of certain other phosphate and thiophosphate-containing materials including esters of carbohydrates, nucleotides and lipids.
Currently, phosphoproteins are most often detected by autoradiography after incorporation of 32P or 33P into cultured cells or after incorporation into subcellular fractions by protein kinases. See, e.g., Yan et al., “Protein Phosphorylation: Technologies for the Identification of Phosphoamino Acids,” J. Chromatogr. A. 808:23-41 (1998); Guy, G., Phillip, R. and Tan, Y. Electrophoresis 15:417-440 (1994). Such approaches are restricted to a limited range of biological materials, such as tissue culture samples and analysis of clinical samples would require in vivo labeling of patients, which is not feasible. Several alternatives to radiolabeling have also been developed over the years.
Phosphoproteins can also be detected by immunoblotting and immunoprecipitation. See, e.g., Soskic et al., “Functional Proteomics Analysis of Signal Transduction Pathways of the Platelet-Derived Growth Factor Beta Receptor,” Biochemistry 38:1757-1764 (1999); Watty et al., “The In Vitro and In Vivo Phosphotyrosine Map of Activated MuSK,” Proc. Natl. Acad. Sci. USA. 97:4585-4590 (2000). Immunoblotting is best performed after blocking unoccupied sites on the solid-phase support with protein solutions, which interferes with microchemical analysis. Removal of the antibody and stain require relatively harsh treatments (i.e., heating to 65° C., incubation with 0.1% SDS and 1 mM DTT). This also poses problems with subsequent use of the protein for sequencing and mass spectrometry. For immunoprecipitation, only the anti-phosphotyrosine antibodies display binding that is tight enough to allow effective isolation. Though high-quality antibodies to phosphotyrosine are commercially available, antibodies that specifically recognize phosphoserine and phosphothreonine residues have been more problematic, often being sensitive to amino acid sequence context. The reliability of these antibodies has been questioned because of potential steric hindrances between the interaction of these antibodies and the phosphoproteins. Moreover, when phosphoproteins are not enriched prior to detection with the antibody, the presence of unrelated proteins co-migrating with the protein of interest may lead to false positive signals. Therefore, identification of phosphorylated proteins using immunoblotting and immunoprecipitation techniques is effectively limited to proteins containing phosphorylated tyrosine residues. See McLachlin & Chait, supra.
Alternatively, phosphorylated proteins can be identified using chromogenic dyes, but with limited success. The cationic carbocyanine dye “Stains-All” (1-ethyl-2-[3-(3-ethylnaphtho[1,2d]thiazolin-2-ylidene)-2-methylpropenyl]-naphtho[1,2d]thiazolium bromide) stains RNA, DNA, phosphoproteins and calcium-binding proteins blue while unphosphorylated proteins are stained red. See, e.g., Green et al., “Differential Staining of Phosphoproteins on Polyacrylamide Gels with a Cationic Carbocyanine Dye,” Anal. Biochem. 56:43-51 (1973); Hegenauer et al., “Staining Acidic Phosphoproteins (Phosvitin) in Electrophoretic Gels,” Anal. Biochem. 78:308-311 (1977); Debruyne, “Staining of Alkali-Labile Phosphoproteins and Alkaline Phosphatases on Polyacrylamide Gels,” Anal. Biochem. 133:110-115 (1983); “Staining of phosphoproteins in polyacrylamide gels with acridine orange”, Seikagaku 45:327-35 (1973). Stains-All is not routinely used to detect phosphoproteins due to poor specificity and low sensitivity. Stains-All is at least 10 times less sensitive than Coomassie Brilliant Blue as a general protein stain and several orders of magnitude less sensitive than 32P-autoradiography or the techniques described in this patent. Another chromogenic method, the GelCode™ Phosphoprotein detection kit (Pierce Chemical Company, Rockford, Ill.), is designed to detect phosphoproteins in gels; however, this method has many limitations. According to this method, phosphoproteins are detected in gels through alkaline hydrolysis of phosphate esters of serine or threonine, precipitation of the released inorganic phosphate with calcium ions, formation of an insoluble phosphomolybdate complex and then visualization of the complex with a dye such as methyl green, malachite green or rhodamine B [as described in Cutting and Roth (1973)]. The detection sensitivity of the staining method is considerably poorer than Coomassie Blue staining, with 80-160 ng of phosvitin, a protein containing roughly 100 phosphoserine residues, being detectable by the commercialized kit. The staining procedure is quite complex (involving seven different reagents) and alkaline hydrolysis requires heating gels to 65° C., which causes the gel matrix to hydrolyze and swell considerably. Since phosphotyrosine residues are not hydrolyzed by the alkaline treatment, proteins phosphorylated at this amino acid residue escape detection by the method. Dyes for the phosphate-selective fluorescence labeling in which a BODIPY dye is covalently attached to a reactive imidazole group has been developed for the detection of pepsin phosphorylation. See, U.S. Pat. No. 5,512,486; Wang & Giese, “Phosphate-Specific Fluorescence Labeling of Pepsin by BO-IMI,” Anal. Biochem. 230:329-332 (1995).
In addition to detecting phosphoproteins, two methods for the chemical derivatization and enrichment of phosphopeptides resulting in isolation of phosphopeptides from complex mixtures exist. See, e.g. Goshe et al., “Phosphoprotein Isotope-Coded Affinity Tag Approach For Isolating and Quantitating Phosphopeptides in Proteome-Wide Analyses,” Anal. Chem. 73:2578-2586 (2001). The first method involves oxidation of cysteine residues with performic acid, alkaline hydrolysis to induce β-elimination of phosphate groups from phosphoserine and phosphothreonine residues, addition of ethanedithiol, coupling of the resulting free sulfhydryl residues with biotin, purification of phosphoproteins by avidin affinity chromatography, proteolytic digestion of the eluted phosphoproteins, a second round of avidin purification and then analysis by mass spectrometry (Oda, Y., Nagasu, T., and Chait, B. Nature Biotechnol. 19:379 (2001)). The first method uses β-elimination to remove phosphate groups that are replaced with a tag, as exemplified with biotinylated thiol groups wherein the peptides could then be isolated by chromatography on avidin resins. An alternative method requires proteolytic digestion of the sample, reduction and alkylation of cysteine residues, N-terminal and C-terminal protection of the peptides, formation of phosphoramidate adducts at phosphorylated residues by carbodiimide condensation with cystamine, capture of the phosphopeptides on glass beads coupled to iodoacetate, elution with trifluoroacetic acid and evaluation by mass spectrometry (Zhou et al., “A Systematic Approach to the Analysis of Protein Phosphorylation,” Nat. Biotechnol. 19:375-378 (2001). These methods are time consuming, require purified phosphopeptides, and are limiting in what can be isolated. Both procedures identified the monophosphorylated trypsin peptide fragment from the test protein β-casein, but both failed to detect the tetraphosphorylated peptide fragment.
Alternatively, a method for combining chemical modification and affinity purification has been shown for the characterization of serine and threonine phosphopeptides in proteins based on the conversion of phosphoserine and phosphothreonine residues to S-(2-mercaptoethyl)cysteinyl or β-methyl-S-(2-mercaptoethyl)cysteinyl residues by β-elimination/1,2-ethanedithiol addition, followed by reversible biotinylation of the modified proteins. After trypsin digestion, the biotinylated peptides are affinity-isolated and enriched, followed by their subsequent structural characterization by liquid chromatography/tandem mass spectrometry (LC/MS/MS). See Adamczyk et al., “Selective Analysis of Phosphopeptides Within a Protein Mixture by Chemical Modification, Reversible Biotinylation and Mass Spectrometry,” Rapid. Commun. Mass Spectrom. 15:1481-1488 (2001).
Fluorescence detection methods appear to offer the best solution to global protein quantitation in proteomics. However, currently, there is no satisfactory method for the specific and reversible fluorescent detection of gel-separated phosphoproteins from complex samples. Derivatization and fluorophore labeling of phosphoserine residues by blocking free sulfhydryl groups with iodoacetate or performate, alkaline β-elimination of the phosphate residue, addition of ethanedithiol, and reaction of the resulting free sulfhydryl group with 6-iodoacetamidofluorescein has been demonstrated in capillary electrophoresis using laser-induced fluorescence detection and similar reactions have been performed on protein microsequencing membranes. However, neither method has been shown to be suitable for detection of phosphoproteins directly in gels. One problem with the approach is that a delicate balance must be struck between the base and the ethanedithiol in order to achieve elimination of the phosphate group from the serine residue and addition of the ethanedithiol to the resulting dehydroalanine residue without hydrolysis of the peptide backbone.
Several instrument-based methods are also available for the determination of protein phosphorylation such as 31P-NMR, mass spectrometry [See, e.g., Resing & Ahn, “Protein Phosphorylation Analysis by Electrospray Ionization-Mass Spectrometery,” Methods Enzymol. 283:29-44 (1997); Aebersold and Goodlett, “Mass Spectrometry in Proteomics,” Chem. Rev. 101:269-295 (2001). Affolter, M., Watts, J., Krebs, D., and Aebersold, R. Anal. Biochem. 223:74 (1994); Liao, P., Leykam, J., Andrews, P., Gage, D., and Allison, J. Anal. Biochem. 219:9 (1994); Oda, Y., Huang, K., Cross, F., Cowburn, D., and Chait, B. Proc. Natl. Acad. Sci. USA 96:6591 (1999)) and protein sequencing. Mass spectrometry has been used to provide the molecular mass of an intact phosphorylated protein by comparing the mass of the unphosphorylated protein to that of the phosphorylated protein. See, e.g., McLachlin & Chait, “Analysis of Phosphorylated Proteins and Peptides by Mass Spectrometry,” Current Opin. Chem. Biol. 5:591-602 (2001). This is limiting in that researchers must have purified amounts of both proteins. While these procedures accurately characterize the phosphorylation status of proteins and peptides, they are unsuitable for high-throughput screening of phosphorylated substrates. The techniques are generally used after a phosphoprotein has been identified by autoradiography or immunoblotting with an anti-phosphotyrosine antibody. Though methods have recently been introduced to directly quantify the relative abundance of phosphoproteins in two different samples by mass spectrometry through culturing different cell populations in 15N-enriched and 14N-enriched medium, the linear dynamic range of such methods has explicitly been demonstrated over only a 10-fold range. Ion suppression phenomena associated with mass spectrometry prevents stoichiometric comparison of different phosphoproteins by such techniques.
For analysis of the site(s) of phosphorylation on molecules, a more detailed analysis of the sites of phosphate attachment and stoichiometery often requires the examination of peptide fragments of the phosphoprotein of interest. Such fragments are usually generated by digestion of the phosphoprotein with proteases such as trypsin. However, mass spectroscopic analysis of proteolytic digests of proteins rarely provides full coverage of the protein sequence and regions of interests often go undetected. In addition, protein phosphorylation is often sub-stoichiometric, such that the phosphoproteins are present in lower abundance than other peptides from the protein of interest. Therefore, the identification and characterization of phosphoproteins would be improved greatly by highly selective methods of enriching phosphopeptides prior to analysis with mass spectrometry. It would be particularly useful to detect phosphoproteins by reagents that do not chemically alter the structure or mass of the phosphoproteins.
Currently, selective enrichment of phosphopeptides from complex mixtures is performed using immobilized metal affinity chromatography, known as IMAC. Using this technique, metal ions such as Fe3+ or Ga+3 are bound to a chelating support prior to the addition of a complex mixture of peptides or proteins. See, e.g., Posewitz & Tempst, “Immobilized Gallium(III) Affinity Chromatography of Phosphopeptides,” Anal. Biochem. 71:2883-2892 (1999). Phosphopeptides that bind to the column can be released using high pH or phosphate buffer, though the latter step usually requires a further desalting step before analysis with mass spectrometry. Resins with iminodiacetic acid and nitrilotriacetic acid chelators are known and are available commercially. See, e.g., Neville et al., “Evidence for Phosphorylation of Serine 753 in CFTR Using a Novel Metal-Ion Affinity Resin and Matrix-Assisted Laser Desorption Mass Spectrometry,” Protein Sci. 6:2436-2445 (1997). However, there are several complications using current techniques, including loss of phosphopeptides that do not bind to the column (low affinity), difficulty in the subsequent elution of phosphorylated peptides, and background from non-phosphorylated peptides that have affinity for immobilized metal ions (low specificity).
Mass spectrometry-based detection of separated peptides and direct matrix-assisted laser desorption/ionization (MALDI) analysis of phosphopeptides bound to an IMAC support has been demonstrated. See Zhou et al., “Detection and Sequencing of Phosphopeptides Affinity Bound to Immobilized Metal Ion Beads by Matrix-Assisted Laser Desorption/Ionization Mass Spectrometry,” J. Am. Soc. Mass. Spectrom. 11:273-283 (2000). IMAC has also been coupled directly to mass spectrometry instruments on-line, or with superseding separation techniques, such as HPLC and capillary electrophoresis (CE), for the detection and analysis of phosphopeptides.
The present invention overcomes the limitations of the current methods by utilizing a cationic transition metal and a compound that comprises a metal-chelating moiety and a chemical moiety, typically a reactive group or label such as a fluorophore, or a combination thereof, to detect phosphoproteins and phosphopeptides. There are a variety of chelating moieties that use poly-carboxylate binding sites to selectively bind monovalent and divalent metal cations, and these are often used in fluorescent calcium ion indicators. Examples of these indicators are, for example, quin-2, fura-2, indo-1 (U.S. Pat. No. 4,603,209); fluo-3 and rhod-2 (U.S. Pat. No. 5,049,673), and FURA RED™ (U.S. Pat. No. 4,849,362). A family of BAPTA-based indicators that are selective for calcium ions are described in HAUGLAND, HANDBOOK OF FLUORESCENT PROBES AND RESEARCH CHEMICALS (9th edition, CD-ROM, September 2002). Examples of BAPTA-based metal-chelators are also described in U.S. Pat. Nos. 5,773,227; 5,453,517; 5,516,911; 5,501,980; 6,162,931 and 5,459,276.
Indicators of free calcium concentrations are based upon selective calcium binding to fluorescent dyes. Though it is well known that BAPTA compounds bind certain divalent cations, such as calcium, as analogs of the common EGTA chelator, the BAPTA compounds are also known to bind almost all inorganic polyvalent cations with an affinity that may be higher or lower than that of the compound for calcium ions. Their selectivity and utility for measuring calcium in biological cells results from the general absence or low abundance of these other polycations in living systems. The affinity and selectivity of the BAPTA-based indicators for polycations, including gallium and similar metals of utility to this invention, can be modified by shifts in pH, solvent composition, ionic strength and other experimental variables. This shift in cation selectivity and affinity is critical to all aspects of the disclosed invention, including both detection and isolation of phosphorylated targets.
The present invention overcomes the limitations and disadvantages of currently disclosed methods for identifying, isolating, analyzing and quantitating phosphorylated proteins and thus provide methods, compounds and compositions to alleviate long-felt needs for rapid and effective high-throughput methods for detecting and isolating phosphoproteins for further analysis. The present invention can accurately identify phosphopeptides and phosphoproteins comprising as few as one phosphate group and in a simple method that does not require multiple steps or pre-treatment of the sample. Importantly, the present invention is the first known method to provide a means for accurately identifying the phosphorylated proteome and allows for the quantitative identification of increased phosphorylation of proteins. The present invention is an important tool for identifying novel phosphorylated proteins in the proteome. The technology has unsurpassed quantitative characteristics, particularly when used in combination with reagents for the detection of total proteins. In addition, as will be described below, the materials and methods of the present invention are not limited to the detection and/or separation of phosphorylated proteins.